Article

Efficacy of Continuation ECT and Antidepressant Drugs

Compared to Long-Term Antidepressants Alone
in Depressed Patients

Gerard G. Gagné, Jr., M.D.
Martin J. Furman, M.D.
Linda L. Carpenter, M.D.
Lawrence H. Price, M.D.
Objective: The purpose of this study was
to evaluate the efficacy of continuation
ECT in depression.
Method: The authors used retrospective
chart review to identify 29 patients who
received continuation ECT plus long-term
antidepressant treatment after a positive
response to acute treatment with ECT for
a depressive episode (continuation ECT
group). A retrospective case-controlled
approach was used to ascertain a match-
ing group of 29 patients who received
long-term antidepressant treatment
alone after responding positively to acute
ECT (antidepressant-alone group). All 58
patients (46 with unipolar depression, 12
with bipolar disorder) had been chroni-
cally depressed before receiving acute
ECT. Data from medical records were ana-
lyzed by using survival analysis and pro-
portional hazards regression to deter-
mine outcome and risk factors.
Results: The mean duration of the fol-
low-up period for all patients was 3.9
years (5.4 years for the continuation ECT
patients and 2.4 years for the antidepres-
sant-alone patients). Outcome was signifi-
cantly better in the continuation ECT
group. The cumulative probability of sur-
viving without relapse or recurrence at 2
years was 93% for continuation ECT pa-
tients and 52% for antidepressant-alone
patients. At 5 years, survival declined to
73% for continuation ECT patients, but fell
to 18% for antidepressant-alone patients.
Mean survival times were 6.9 years for the
continuation ECT patients and 2.7 years
for the antidepressant-alone patients.
Conclusions: T h e f i n d i n g s p r o v i d e
strong support for the efficacy of continu-
ation ECT plus long-term antidepressant
treatment in preventing relapse and re-
currence in chronically depressed pa-
tients who have responded to acute treat-
ment with ECT.

(Am J Psychiatry 2000; 157:1960–1965)

T he efficacy of ECT is well established for the acute

treatment of depression, particularly in patients with very
severe symptoms, mood-congruent delusions, prominent
suicidality, and inanition and dehydration (1–5). Unfortu-
nately, this population of severely ill patients remains at
high risk for relapse and recurrence, even when adequate
extended pharmacotherapy is provided (6). The identifi-
cation of effective continuation and maintenance strate-
gies for these patients is a critical issue in the long-term
management of depression.
Continuation therapy is often narrowly defined as
treatment that occurs during the several months immedi-
ately after resolution of an acute episode of illness, with
the primary purpose of relapse prevention (7). Mainte-
nance therapy is sometimes differentiated as a more pro-
longed course of intervention, extending beyond contin-
uation therapy and used to prevent recurrence of the
illness (i.e., a new episode). In practice, the distinction be-
tween continuation therapy and maintenance therapy is
arbitrary (5); what is important is that continuation ther-
apy of some sort is necessary to prevent the return of de-
pressive symptoms, irrespective of whether ECT or anti-
depressant medications are used to treat an acute episode
of depression (8).
Although the practice of extended and long-term drug
therapy for depression has become standard, the use of
broadly defined continuation ECT has failed to gain gen-
eral acceptance (9). Proponents of continuation ECT ar-
gue that this approach represents an effective alternative
or adjunct to drug treatment in selected patients who are
at high risk of relapse. This group includes patients who
have demonstrated response of their acute episode to a
standard course of ECT and 1) have experienced an acute
episode that was refractory to drugs, 2) have relapsed dur-
ing long-term pharmacological treatment, or 3) have had
contraindications to long-term antidepressant or thymo-
leptic drug treatment (7, 10, 11). The published literature
on continuation ECT consists mainly of case reports (12–
15), naturalistic studies (8, 16), and retrospective case se-
ries (11, 17–20) of patients who clearly obtained benefit
from continuation ECT. We are aware of no prospective or
randomized studies, and only three reports have included
some form of controlled comparison, even if retrospective
(11, 16, 18).

1960 Am J Psychiatry 157:12, December 2000

 GAGNÉ, FURMAN, CARPENTER, ET AL.

TABLE 1. Demographic and Clinical Characteristics of Patients Who Responded to Acute Treatment With ECT and Then Re-
ceived Either Continuation ECT Plus Long-Term Antidepressant Treatment or Long-Term Antidepressant Treatment Alone
Patients Receiving Comparison
Variable Continuation ECT (N=29) Group (N=29) Analysis
N % N % χ2 df p

Female gender 26 89.7 24 82.8 0.58 1 n.s.

Caucasian ethnicity 29 100.0 28 96.6 1.02 1 n.s.
Employment 0.71 3 n.s.
None 6 20.7 3 10.3
Employed 5 17.2 6 20.7
Retired 13 44.8 13 44.8
Disabled 5 17.2 7 24.1
Marital status 1.29 4 n.s.
Never married 5 17.2 4 13.8
Married/cohabitant 14 48.3 15 51.7
Separated 0 0.0 1 3.4
Divorced 4 13.8 3 10.3
Widowed 6 20.7 6 20.7
Family history of depression 11 37.9 17 58.6 2.49 1 n.s.
Primary axis I diagnosis 1.93 1 n.s.
Major depression (unipolar) 24 82.8 22 75.9
Bipolar I/II (depressed) 5 17.2 7 24.1
Number of secondary axis I diagnoses 1.89 2 n.s.
None 17 58.6 18 62.1
One 9 31.0 6 20.7
Two or more 3 10.3 5 17.2
ECT electrode placement 0.09 2 n.s.
Unilateral 8 27.6 7 24.1
Bilateral 19 65.5 20 69.0
Crossover 2 6.9 2 6.9

Mean SD Mean SD t df p
Age (years)
At the time of the study 64.5 19.0 66.1 18.7 0.32 55 n.s.
At onset of depression 40.2 18.0 50.8 19.4 1.96 46 n.s.
At first ECT 51.0 18.2 59.9 17.2 1.40 30 n.s.
At index ECT 57.5 17.5 67.8 17.4 1.33 33 n.s.
Number of adequate antidepressant trials before index ECT 4.1 3.8 2.2 2.2 2.35 56 <0.03
Number of treatments during acute ECT 6.0 2.8 6.8 3.0 1.05 56 n.s.

The present retrospective case-controlled study was un-
dertaken to evaluate the efficacy of continuation ECT plus
long-term antidepressant treatment versus long-term an-
tidepressant treatment alone in a large sample of patients
with severe chronic depression treated under naturalistic
conditions. Before beginning long-term treatment of any
sort, all patients had responded positively to a standard
course of ECT given for an acute episode of illness. Both
within- and between-subjects comparisons were used,
and outcome measures corresponded to those used in
previous reports (12). Clinical and demographic factors
were evaluated with respect to their ability to predict out-
come. On the basis of the published literature, we hypoth-
esized that continuation ECT combined with long-term
antidepressant treatment would be superior to standard
treatment with long-term antidepressants alone in these
severely ill patients.
Method
Patients
To determine patients’ eligibility for the study, we reviewed the
charts of all patients who received continuation ECT from April
1985 to July 1999 in the ECT Program at Butler Hospital, a univer-
sity-affiliated psychiatric facility in Providence, R.I. Continuation
ECT patients with a primary DSM-III, DSM-III-R, or DSM-IV axis
I diagnosis of either major depression (unipolar) or bipolar I or bi-
polar II disorder (most recent episode depressed) were enrolled
(continuation ECT group). Excluded were continuation ECT pa-
tients who 1) had severe concurrent medical illness, 2) had a pre-
existing nonaffective psychiatric illness, or 3) met criteria for
schizoaffective disorder. This protocol was reviewed by the Butler
Hospital institutional review board and judged to be exempt from
the requirement for written informed consent.
A comparison group consisting of patients given long-term an-
tidepressant treatment alone (antidepressant-alone group) was
matched to the continuation ECT group for age, gender, primary
axis I diagnosis, age at onset of depression, presence of comorbid
psychosis, and year during which the index course of acute ECT
was administered. Like the continuation ECT group, all patients
in the antidepressant-alone group had had a positive response to
acute treatment with ECT. Patients included in the antidepres-
sant-alone group were subjected to the same inclusion/exclusion
criteria as the continuation ECT patients. The patients in the an-
tidepressant-alone group had received four to 15 treatments of
acute ECT for their depressive episode. After acute ECT, these pa-
tients were continued or maintained with antidepressant medi-
cations alone, without additional ECT, for prevention of relapse
and recurrence.
Additional data on demographic and clinical characteristics
were abstracted from all patients’ charts. Additional demographic
characteristics included ethnicity, marital status, and employ-

Am J Psychiatry 157:12, December 2000 1961

CONTINUATION ECT
ment history. Ancillary clinical characteristics included second-
ary axis I diagnoses, any axis II diagnoses, bipolar versus unipolar
subtype, family history of affective disorder in first-degree rela-
tives, age at onset of depressive symptoms, age at first ECT, age at
index ECT, medications administered during the index depressive
episode (including number, duration of trials, and doses of anti-
depressants, neuroleptics, thymoleptics, and anxiolytics), num-
ber and duration of hospitalizations, and the use of psychother-
apy. Clinical information suggestive of greater severity (psychosis,
severe suicidality, and melancholia) (7, 16, 21) was noted. The
method of electrode placement (unilateral versus bilateral) and
mean seizure duration were determined.
A total of 58 patients were included in the study, 29 in the con-
tinuation ECT group and 29 in the antidepressant-alone compar-
ison group. The mean age of the continuation ECT patients at the
time of the study was 64.5 years (SD=19.0, range=29–91) and that
of the antidepressant-alone patients was 66.1 years (SD=18.7,
range=34–96). Both groups were predominantly (>80%) female
and appeared similar in terms of socioeconomic background and
ethnicity (Table 1).
After the course of acute ECT, 28 (96.6%) of 29 continuation
ECT patients and all of the antidepressant-alone patients were
managed with antidepressant medications. One continuation
ECT patient received no psychotropic medication during the
course of continuation ECT. A comparably broad range of agents
was used in both groups, including tricyclics (amitriptyline, clo-
mipramine, desipramine, doxepin, imipramine, nortriptyline),
heterocyclics (amoxapine, bupropion, venlafaxine), monoamine
oxidase inhibitors (phenelzine, tranylcypromine), selective sero-
tonin reuptake inhibitors (sertraline, fluoxetine, paroxetine), and
serotonin receptor antagonists (nefazodone, trazodone). Patients
with bipolar disorder in both groups received mood stabilizers
(carbamazepine, lithium, divalproex), and a similar proportion of
psychotic patients in both groups received neuroleptics. No pa-
tients with bipolar disorder were maintained with antidepres-
sants without a mood stabilizer. Drug doses were generally in ac-
cepted therapeutic ranges and were similar between treatment
groups.
ECT Treatment
ECT was administered by using a Mecta SR-1 (Mecta, Portland,
Ore.) brief-pulse, constant-current device. ECT duration was con-
sidered optimal if the seizure duration (determined by observable
tonic-clonic seizure) lasted 30–60 seconds. All ECT was given by a
single psychiatrist (M.J.F.), who adjusted the treatment frequency
on the basis of the patient’s clinical response. Patients generally
received ECT for acute episodes at a rate of two to three treat-
ments/week until a positive clinical response occurred, on the
basis of the consensus evaluation of M.J.F. and the referring psy-
chiatrist that the core depressive symptoms necessitating hospi-
tal admission were much or very much improved. The decision to
institute continuation ECT was similarly based on the clinical
consensus of M.J.F. and the referring psychiatrist that medication
maintenance alone was unlikely to be effective given the patient’s
history of treatment response and most recent clinical presenta-
tion. If the patient continued with ECT, treatments generally were
delivered weekly for the first month, every 2 weeks for the follow-
ing month, and then monthly.
Study Design and Methods
This retrospective study evaluated the efficacy of continuation
ECT by comparing the outcomes of patients who received this
treatment combined with long-term antidepressant treatment
(continuation ECT group) with those of matched comparison pa-
tients who received long-term antidepressant treatment alone
(antidepressant-alone group). Both groups had manifested a pos-
itive response to acute treatment with ECT just before entering
1962
the period of long-term treatment under study. In addition, pa-
tients in the continuation ECT group were included in a mirror-
image comparison of their own clinical course during the time
period immediately preceding the index episode of illness versus
their course afterward.
Acute ECT was defined as a standard course of ECT adminis-
tered for the acute treatment of an index episode of depression.
The index ECT was the first treatment in the course of acute ECT,
and it served as the reference point for calculating the major out-
come variables and for the beginning of the survival analysis.
Continuation ECT was defined as ongoing ECT that was adminis-
tered after completion of the course of thrice weekly acute ECT,
beginning with the administration of ECT on a weekly or less fre-
quent basis.
The four outcome measures of primary interest were the prob-
ability of surviving without relapse or recurrence, the time to re-
lapse/recurrence, the frequency of hospitalization quotient, and
the hospital day quotient. Probability of survival and time to re-
lapse/recurrence were evaluated by using survival analysis, with
relapse/recurrence stringently defined as the reemergence of de-
pressive symptoms of sufficient severity to result in either rehos-
pitalization or a new course of acute ECT. Time was counted in
years from the date of the index ECT. Data from patients who did
not experience relapse or recurrence during the observation pe-
riod were censored. The mean duration of the observation period
for all patients was 3.9 years (SD=3.1) (for continuation ECT pa-
tients, mean=5.4 years, SD=3.1; for antidepressant-alone pa-
tients, mean=2.4 years, SD=2.4).
As described elsewhere (12), the frequency of hospitalization
quotient and hospital day quotient were used to quantify the clin-
ical history and functional status of each patient. Each variable
was computed for the observation period after index ECT and for
an equivalent time period preceding the index ECT. The fre-
quency of hospitalization quotient, reflecting the average num-
ber of hospital admissions per year, was calculated by dividing
the total number of admissions by the number of days in the ob-
servation period and multiplying the quotient by 365. The hospi-
tal day quotient, reflecting the average number of psychiatric in-
patient hospital days, was determined in a similar manner. Both
measures were calculated only for patients with an observation
period ≥6 months (periods of <6 months would have resulted in
artificially inflated quotients).
Statistical Analysis
Comparisons of clinical and demographic characteristics be-
tween groups were performed with t tests for continuous vari-
ables or contingency tables (chi-square or Fisher’s exact test) for
categorical variables. The Kaplan-Meier product-limit method
was used to estimate survival curves. Differences between the
continuation ECT and antidepressant-alone groups were evalu-
ated by using the Mantel-Cox test. The Cox proportional hazards
regression model was used to determine the significance of po-
tential risk factors for relapse/recurrence. All statistical tests were
two-tailed, with significance set at p<0.05. Analyses were per-
formed with SPSS, version 6.1 (22).
Results
Patient and Treatment Characteristics
Table 1 presents selected clinical and demographic
characteristics of the 58 patients involved in the study. In
general, patients in the continuation ECT and antidepres-
sant-alone groups appeared very similar in clinical and
demographic features. The single significant difference
between groups was the mean number of adequate medi-
Am J Psychiatry 157:12, December 2000

cation trials before the index ECT. (“Adequacy” was de-
fined as at least 4 weeks of treatment at a dose equal to the
manufacturer’s recommended minimum effective dose).
Patients in the continuation ECT group had received a
mean of 4.1 trials (SD=3.7), whereas the antidepressant-
alone patients had received 2.2 trials (SD=2.2) (t=2.36, df=
56, p<0.03).
The treatment parameters of the acute ECT given to the
continuation ECT and the antidepressant-alone patients
were also similar (Table 1). About two-thirds of the pa-
tients in each group received bilateral treatment. The
mean number of treatments during acute ECT did not dif-
fer significantly between groups. The mean seizure dura-
tion during the course of acute ECT was nearly identical
for both groups, with a pooled mean of 39.7 seconds (SD=
10.4). Mean seizure duration for the remainder of treat-
ments after the acute course in the continuation ECT pa-
tients was 44.7 seconds (SD=10.5).
Survival Analysis and Risk Factors
Figure 1, depicting the Kaplan-Meier survival curves for
continuation ECT and antidepressant-alone patients,
shows that outcome was superior for the continuation
ECT group. The cumulative probability of surviving with-
out relapse or recurrence at 2 years after index ECT was
93% for continuation ECT patients and 52% for antide-
pressant-alone patients. At 5 years, survival declined to
73% for continuation ECT patients, but fell to 18% for an-
tidepressant-alone patients. Between 5.5 and 7 years, sur-
vival appeared to drop substantially for the continuation
ECT patients, although it remained higher than that for
antidepressant-alone patients. Mean survival time for
continuation ECT patients was 6.9 years (SD=4.4) (95%
confidence interval [CI]=5.3–8.5 years), with a median of
6.6 years (SD=3.9) (95% CI=5.1–8.0 years). In contrast,
mean survival time for antidepressant-alone patients was
2.7 years (SD=2.5) (95% CI=1.7–3.6 years), with a median
of 2.2 years (SD=3.7) (95% CI=0.9–3.6 years).
Potential risk factors for relapse/recurrence, including
secondary axis I diagnoses, axis II diagnoses, bipolar versus
unipolar subtype, concomitant psychosis or melancholia,
prominent suicidality, number of adequate medication tri-
als before index ECT, and adjunctive psychotherapy after
acute ECT, were not statistically significant in analyses us-
ing proportional hazards regression.
Frequency of Hospitalization Quotient
and Hospital Day Quotient
In the continuation ECT group, the frequency of hospi-
talization quotient was 3.4 (SD=6.6) before the index ECT
and 0.8 (SD=1.3) after the index ECT. The hospital day
quotient for this group was 57 (SD=112) before the index
ECT and 9 (SD=17) after the index ECT. In the antidepres-
sant-alone group, the preindex frequency of hospitaliza-
tion quotient was 1.2 (SD=1.2) and the postindex fre-
quency of hospitalization quotient was 1.2 (SD=1.0). The
Am J Psychiatry 157:12, December 2000
GAGNÉ, FURMAN, CARPENTER, ET AL.
FIGURE 1. Time to Relapse or Recurrence of Depression in
Patients Who Responded to Acute Treatment With ECT and
Then Received Either Continuation ECT Plus Long-Term An-
tidepressant Treatment or Long-Term Antidepressant
Treatment Alonea
Number Remaining at Risk
Proportion of Patients Without Relapse
29 25 18 12 5 1 1 0
29 14 8 4 0 0 0 0
1.0
Patients receiving
continuation ECT
0.8
Comparison group
0.6
0.4
0.2
0.0
0 2 4 6 8 10 12 14
Years to Relapse
a Mantel-Cox χ2=19.87, df=1, p<0.00005.
preindex hospital day quotient for this group was 17 (SD=
15), and the postindex hospital day quotient was 13 (SD=
11). There were no statistically significant differences be-
tween the continuation ECT group and antidepressant-
alone group in the frequency of hospitalization quotient
or the hospital day quotient before or after the index ECT
or in the difference between the pre- and postindex quo-
tients. Similarly, there were no significant pre- versus
postindex differences in the frequency of hospitalization
quotient or the hospital day quotient within either group,
although both measures showed declines in the continua-
tion ECT group.
Discussion
In this sample of chronically depressed patients re-
sponding to a standard course of acute ECT, continuation
ECT in conjunction with long-term antidepressant treat-
ment resulted in a marked improvement in clinical out-
come. In comparison with rates of survival for patients
who received long-term antidepressant treatment alone,
rates of survival without relapse or recurrence for patients
who received continuation ECT and long-term antide-
pressant treatment were nearly doubled (93% versus 52%)
at 2 years and showed a fourfold increase (73% versus
18%) by 5 years. The mean time to relapse or recurrence
more than doubled in the continuation ECT group (6.9
years versus 2.7 years for the antidepressant-alone group).
Comparison patients were closely matched with contin-
uation ECT patients on standard demographic variables
(i.e., gender and age), as well as on clinical variables (i.e.,
primary axis I diagnosis, age at onset of depression, the
presence of comorbid psychosis), a treatment variable
(i.e., positive response to an index course of ECT), and a
1963
CONTINUATION ECT
temporal variable (i.e., year during which the index course
of acute ECT was administered). As a result, patients in
both groups appeared similar, as demonstrated by the lack
of significant differences between groups across a broad
array of clinical and treatment variables. The only signifi-
cant difference detected was the greater number of previ-
ous adequate antidepressant trials in the continuation
ECT group, suggesting that the depressive symptoms of
these patients were more refractory to pharmacotherapy.
However, neither this nor any other identified factor was
related to risk of relapse/recurrence.
The follow-up period in the present study lasted more
than 5 years for the continuation ECT patients, comparing
favorably with the results of previous studies of continua-
tion ECT that have employed a comparison group, none of
which extended beyond a 1-year follow-up (11, 16, 18). In
this regard, it is notable that survival rates for the continu-
ation ECT patients appeared to deteriorate appreciably
between 5.5 and 7 years. It is unclear whether this decline
represents a true clinical phenomenon or merely the di-
minished reliability of the right-hand tail of the survival
curve with smaller numbers of at-risk patients (23). For ex-
ample, although the survival curve in Figure 1 suggests
that all patients in both treatment groups eventually re-
lapsed, a marked preponderance of relapses (27 [93.1%] of
29 patients) occurred in the antidepressant-alone group;
11 (37.9%) of the 29 continuation ECT patients were cen-
sored before relapse.
In a previous study that also utilized survival analysis,
Schwarz et al. (18) detected no significant differences be-
tween continuation ECT patients and comparison pa-
tients. The lack of difference may have been due partly to
the relatively short (6-month) follow-up period employed
by those investigators. Surprisingly, hospitalization rates
and number of hospital days, as reflected by the frequency
of hospitalization quotient and hospital day quotient, re-
spectively, proved insensitive to the effects of continua-
tion ECT in between-groups comparisons in the present
study. This finding is in contrast to the significant findings
of other investigators who used these or analogous mea-
sures (3, 16, 21, 24). Although our findings for the mean
frequency of hospitalization quotient and mean hospital
day quotient were similar to those previously reported,
greater variance in our data precluded findings of statisti-
cally significant differences. Significant within-group de-
creases before and after continuation ECT were also not
demonstrable, even though mean changes were in the
predicted direction. Again, these results appeared to be
due to marked variability and artifacts related to the man-
ner in which the frequency of hospitalization quotient and
the hospital day quotient were computed for the time pe-
riod before the index ECT, even though efforts were made
to minimize such artifacts. In any case, these observations
underscore the importance of utilizing the more robust
1964
survival analytic techniques in subsequent research on
continuation ECT.
The present findings must be considered in light of the
many caveats that apply to naturalistic, retrospective
studies. The sample size, although among the largest yet
reported on this topic, was still modest, resulting in lim-
ited power to detect possibly meaningful differences in
the clinical and demographic characteristics of the two
groups. Assignment of patients to treatment groups was
not random, and assessments of outcome were not blind
or prospectively systematic. Diagnostic and clinical data
were not derived from structured instruments and, even
though the medical records were relatively complete, de-
tailed information on patients’ clinical and functional
status over time was not available. The use of hospitaliza-
tion or initiation of a new course of acute ECT as mea-
sures of relapse/recurrence may have been overly strin-
gent, resulting in an underestimation of the true relapse/
recurrence rates. The relatively older age, female prepon-
derance, and Caucasian ethnicity of the study group limit
generalizability, as does the fact that all of the patients in
this study demonstrated a positive response to a standard
course of acute ECT before entering long-term treatment.
Finally, even though treatment groups were matched on
several clinical and demographic variables of obvious rel-
evance, it is possible that other key clinical or treatment
variables that were not subjected to matching or assess-
ment may have differed between groups and may have af-
fected outcome.
It is possible that the more favorable outcome of pa-
tients in the continuation ECT group was a result of nonbi-
ological and nonpharmacological aspects of the different
follow-up treatments. Patients who received continuation
ECT were required to have direct contact with clinicians at
least once a month to maintain treatment. If a continua-
tion ECT patient missed an appointment, the ECT staff at-
tempted to contact and reengage the patient. Such active
efforts to ensure compliance were probably less vigorous
for patients in the antidepressant-alone group. As a result,
continuation ECT patients may have benefited both from
the psychosocial support of additional contacts with treat-
ers and from an increased likelihood that the biological
treatment was actually being taken.
In summary, this study demonstrates that, in chroni-
cally depressed patients who have responded to an acute
course of ECT, continuation ECT in combination with an-
tidepressants is more effective than antidepressants alone
in preventing relapse and recurrence. Prospective re-
search currently underway should provide more definitive
data about the magnitude and duration of this effect, as
well as about relevant prognostic factors (25). In the
meantime, this study supports the continued judicious
use of this therapeutic approach in patients who have
failed to tolerate or benefit from long-term antidepressant
pharmacotherapy.
Am J Psychiatry 157:12, December 2000

Received Dec. 22, 2000; revision received June 13, 2000; accepted
July 12, 2000. From the Mood Disorders Program, Butler Hospital;
and the Department of Psychiatry and Human Behavior, Brown Uni-
versity School of Medicine. Address reprint requests to Dr. Price, But-
ler Hospital, Department of Psychiatry and Human Behavior, Brown
University School of Medicine, 345 Blackstone Blvd., Providence RI
02906; lawrence_price_md@brown.edu (e-mail).
Supported in part by a grant from the Department of Psychiatry
and Human Behavior, Brown University School of Medicine.
The authors thank Jason Siniscalchi, M.S., and Christine A. Richard.
References
1. Martin BA: Electroconvulsive therapy for depression in general
psychiatric practice. Psychiatr J Univ Ottawa 1989; 14:413–417
2. Schatzberg AF: Recent developments in the acute somatic
treatment of major depression. J Clin Psychiatry 1992; 53(Mar
suppl):20–25
3. Persad E: Electroconvulsive therapy in depression. Can J Psychi-
atry 1990; 35:175–182
4. Philibert RA, Richards L, Lynch CF, Winokur G: Effect of ECT on
mortality and clinical outcome in geriatric unipolar depres-
sion. J Clin Psychiatry 1995; 56:390–394
5. Fink M, Abrams R, Bailine S, Jaffe R: Ambulatory electroconvul-
sive therapy: report of a task force of the Association for Con-
vulsive Therapy. Convuls Ther 1996; 12:42–55
6. Grunhaus L, Pande AC, Haskett RF: Full and abbreviated
courses of maintenance electroconvulsive therapy. Convuls
Ther 1990; 6:130–138
7. Rabheru K, Persad E: A review of continuation and mainte-
nance electroconvulsive therapy. Can J Psychiatry 1997; 42:
476–484
8. Clarke TB, Coffey CE, Hoffman GW Jr, Weiner RD: Continuation
therapy for depression using outpatient electroconvulsive
therapy. Convuls Ther 1989; 5:330–337
9. Sackeim HA: Continuation therapy following ECT: directions for
future research. Psychopharmacol Bull 1994; 30:501–521
10. Monroe RR Jr: Maintenance electroconvulsive therapy. Psychi-
atr Clin North Am 1991; 14:947–960
Am J Psychiatry 157:12, December 2000
GAGNÉ, FURMAN, CARPENTER, ET AL.
11. Petrides G, Dhossche D, Fink M, Francis A: Continuation ECT: re-
lapse prevention in affective disorders. Convuls Ther 1994; 10:
189–194
12. Thienhaus OJ, Margletta S, Bennett JA: A study of the clinical ef-
ficacy of maintenance ECT. J Clin Psychiatry 1990; 51:141–144
13. Decina P, Guthrie EB, Sackeim HA, Kahn D, Malitz S: Continua-
tion ECT in the management of relapses of major affective ep-
isodes. Acta Psychiatr Scand 1987; 75:559–562
14. Loo H, Galinowski A, de Carvalho W, Bourdel MC, Poirier MF:
Use of maintenance ECT for elderly depressed patients (letter).
Am J Psychiatry 1991; 148:810
15. Kramer BA: Maintenance electroconvulsive therapy in clinical
practice. Convuls Ther 1990; 6:279–286
16. Aronson TA, Shukla S, Hoff A: Continuation therapy after ECT
for delusional depression: a naturalistic study of prophylactic
treatments and relapse. Convulsive Ther 1987; 3:251–259
17. Stiebel VG: Maintenance electroconvulsive therapy for chroni-
cally mentally ill patients: a case series. Psychiatr Serv 1995;
46:265–268
18. Schwarz T, Loewenstein J, Isenberg KE: Maintenance ECT: indi-
cations and outcome. Convuls Ther 1995; 11:14–23
19. Jaffe R, Dubin W, Shoyer B, Roemer R, Sharon D, Lipschutz L:
Outpatient electroconvulsive therapy: efficacy and safety. Con-
vuls Ther 1990; 6:231–238
20. Thornton JE, Mulsant BH, Dealy R, Reynolds CF III: A retrospec-
tive study of maintenance electroconvulsive therapy in a uni-
versity-based psychiatric practice. Convuls Ther 1990; 6:121–
129
21. Olfson M, Marcus S, Sackiem HA, Thompson J, Pincus HA: Use
of ECT for the inpatient treatment of recurrent major depres-
sion. Am J Psychiatry 1998; 155:22–29
22. SPSS Advanced Statistics User’s Guide 6.1. Chicago, SPSS, 1994
23. Parmar MKB, Machin D: Survival Analysis: A Practical Ap-
proach. Chichester, UK, John Wiley & Sons, 1995
24. Vanelle JM, Loo H, Galinowski A, de Carvalho W, Bourdel MC,
Brochier P, Bouvet O, Brochier T, Olie JP: Maintenance ECT in
intractable manic-depressive disorders. Convuls Ther 1994; 10:
195–205
25. Rudorfer MV, Lebowitz BD: Progress in ECT research (letter).
Am J Psychiatry 1999; 156:975
1965